Py-faster-rcnn: IndexError: Index out of range

[wait1988]

when I run train_faster_rcnn_alt_opt.py,error occurs:

I1210 14:39:47.685863 223288 layer_factory.hpp:76] Creating layer input-data
I1210 14:39:47.786041 223288 net.cpp:112] Creating Layer input-data
I1210 14:39:47.786516 223288 net.cpp:435] input-data -> data
I1210 14:39:47.787017 223288 net.cpp:435] input-data -> im_info
I1210 14:39:47.788019 223288 net.cpp:435] input-data -> gt_boxes
I1210 14:39:47.804545 223288 python_layer.hpp:43] Traceback (most recent call last):
File "F:\sunqin\DeepLearning\Caffe\Projects\faster-rcnn\tools..\lib\roi_data_layer\layer.py", line 125, in setup
top[idx].reshape(1, self._num_classes * 4)
IndexError: Index out of range
Process Process-1:
Traceback (most recent call last):
File "D:\Anaconda\lib\multiprocessing\process.py", line 258, in _bootstrap
self.run()
File "D:\Anaconda\lib\multiprocessing\process.py", line 114, in run
self._target(_self._args, *_self._kwargs)
File "F:\sunqin\DeepLearning\Caffe\Projects\faster-rcnn\toolstrain_faster_rcnn_alt_opt.py", line 130, in train_rpn
max_iters=max_iters)
File "F:\sunqin\DeepLearning\Caffe\Projects\faster-rcnn\tools..\lib\fast_rcnntrain.py", line 134, in train_net
pretrained_model=pretrained_model)
File "F:\sunqin\DeepLearning\Caffe\Projects\faster-rcnn\tools..\lib\fast_rcnntrain.py", line 43, in __init__
self.solver = caffe.SGDSolver(solver_prototxt)
File "F:\sunqin\DeepLearning\Caffe\Projects\faster-rcnn\tools..\lib\roi_data_layer\layer.py", line 125, in setup
top[idx].reshape(1, self._num_classes * 4)
IndexError: Index out of range

Anyone knows how to fix it?

[stevenxuyan]

I met the same problem, is this problem solved? thanks.

[hoticevijay]

I tried training the VOC dataset without weights using end2end algorithm. I am also facing the same issue. The value of idx is 3 and it seems like top[3] doesn't exist .

[chrert]

Do you provide a config file (eg. experiments/cfgs/faster_rcnn_end2end.yml)? Looks like cfg.TRAIN.HAS_RPN is false but it should be true! Please have a look at experiments/scripts/faster_rcnn_end2end.sh for details.

[hoticevijay]

The config file is present and has TRAIN.HAS_RPN = True. Also the experiments/scripts/faster_rcnn_end2end.sh is correctly pointing to that file. I am getting this issue while training with ZF network. The VGG16 network trains without any issue.

[chrert]

Do you get the error in the same line as wait1988? If you look at the code of lib\roi_data_layer\layer.py you will see that line 125 clearly belongs to the else branch of if cfg.TRAIN.HAS_RPN, so I still suppose the config isn't right. Maybe some debugging could help?

[Johannes-brahms]

any one fix this problem ?
i got the same error when i train on custom data with vgg16 ...

[CWOA]

@Johannes-brahms Did you ever find a fix? Like you, I need to train on custom data with VGG16 and am getting this error..

[wait1988]

I use end2end training and it works for ZF and VGG16 model

[CWOA]

I'm similarly using end2end and VGG16 and getting the error you originally posted.

[limorbagizada]

I tried training the VOC dataset using end2end algorithm and face the same issue.
I tried to give the train_net this file as a parameter: --cfg /home/limor/FRCN_ROOT/py-faster-rcnn/experiments/cfgs/faster_rcnn_end2end.yml but the file is loaded to cfg_from_file variable and the algorithm ignore it since it read from cfg variable and not from cfg_from_file (see file /home/limor/FRCN_ROOT/py-faster-rcnn/lib/fast_rcnn/train.py)
It work if I hardcoded change the /home/limor/FRCN_ROOT/py-faster-rcnn/lib/fast_rcnn/config.py following the yml file. but changing harcoded should not be the solution...
what should need to do ?

[vj-1988]

@chrert Thank you. Your solution worked!

[Mengmengbai]

delete this file py-faster-rcnn/data/cache/ voc_2007_trainval_gt_roidb.pkl

[jamessmith123456]

Do you provide a config file (eg. experiments/cfgs/faster_rcnn_end2end.yml)? Looks like cfg.TRAIN.HAS_RPN is false but it should be true! Please have a look at experiments/scripts/faster_rcnn_end2end.sh for details.

Nice solution!Thanks!